# The Demographic Regulatory Doom Loop: When Safe Asset Rules Meet Aging Fiscal Reality

## Abstract

We document a self-reinforcing loop between population aging, pension obligations, fiscal deterioration, and sovereign risk. Using a panel of 31 sovereign-rated issuers (1990-2024), we show that: (1) old-age dependency drives government expenditure 6:1 faster than revenue, with pension spending absorbing the bulk of the asymmetry; (2) pension spending interacts with demographics to accelerate debt accumulation (pension x OADR interaction p=0.04); (3) the expenditure-revenue gap interacts with aging to compound debt dynamics (interaction p<0.001); and (4) the OADR x debt interaction marginally predicts sovereign rating deterioration (p=0.07). Japan, France, Italy, Finland, and Belgium score highest on our composite doom loop index. The mechanism is recursive: aging electorates demand pension generosity, which widens the expenditure-revenue gap, which accelerates debt accumulation in precisely the countries where demographic pressure is strongest. Basel III and Solvency II zero risk weights for sovereign debt create a regulatory overlay that forces pension funds and insurers to hold the very bonds whose fiscal backing is eroding. This paper quantifies the cross-sectional evidence for the doom loop and projects its intensification through 2050.

---

**JEL Classification:** H55, H63, J11, G28, E62

**Keywords:** demographics, pension spending, sovereign risk, doom loop, fiscal sustainability, sovereign-bank nexus, safe assets, regulatory capital

## 1. Introduction

The interaction between population aging and fiscal sustainability has been extensively studied (Bohn, 1998; De Nardi et al., 2016; Peters, 2026 fiscal dominance paper). The expenditure-revenue asymmetry is well-documented: a 10 percentage point increase in old-age dependency raises government expenditure by 13 percentage points of GDP but revenue by only 4 percentage points (Peters, 2026). What has received less attention is the *recursive* nature of this relationship and its interaction with financial regulation.

We identify a demographic regulatory doom loop with four components:

1. **Aging drives pension generosity**: Old-age dependency predicts pension spending/GDP with coefficient 50.7 (p<0.001), but does not predict revenue (p=0.45). The expenditure-revenue gap widens mechanistically through the pension channel.

2. **Pension spending accelerates debt in aging economies**: The pension-OADR interaction on debt accumulation is positive and significant (p=0.04). Pension spending alone does not predict debt changes (p=0.44), but the interaction with aging does — the fiscal cost of pensions is disproportionately concentrated in already-aging economies.

3. **Debt accumulation degrades sovereign ratings**: The OADR x debt interaction on sovereign ratings is marginally negative (p=0.07), consistent with the safe asset cliff paper's finding of a spline at 20% OADR. High debt is tolerable in young economies but not in aging ones.

4. **Regulatory rules force holding of degraded sovereigns**: Basel III risk weights (0% for AA-rated sovereigns), Solvency II capital charges, and pension fund allocation mandates force institutional investors to hold sovereign debt from the very issuers undergoing this loop. This creates a regulatory lock-in that prevents market discipline from operating until the rating cliff is reached.

The doom loop is recursive because components 1-3 feed back on each other: more aging → more pension spending → more debt → worse ratings → eventually, regulatory reclassification → forced selling → crisis. The regulatory overlay (component 4) suppresses market signals until the cliff is reached, making the eventual adjustment more abrupt.

### 1.1 Relationship to the Research Program

This paper builds on several companion papers in the Perspective on Risk Working Paper Series:

- The **fiscal dominance paper** (Paper 7) documents the 3.3:1 expenditure-revenue asymmetry and identifies the pension channel as the dominant mechanism (80% pensions/transfers, 20% healthcare).
- The **safe asset cliff paper** (Paper 16) documents the OADR spline at 20% (-17.49***) and projects 24 safe issuers declining to 13.5 by 2054.
- The **fragility paper** (Paper 15) confirms fiscal results as the most robust in the portfolio — within-country mechanisms survive panel expansion.
- The **capstone paper** (Paper EE) identifies income as the master moderator, with the safe-issuer subsample showing the strongest rate compression effects.

This paper connects these findings through the regulatory lens: the same demographic forces that make fiscal dynamics robust also make the regulatory doom loop increasingly dangerous.

---

## 2. Data and Methods

### 2.1 Sample

We use the sovereign-rated issuer panel from the safe asset cliff paper: 31 countries with S&P sovereign ratings from 1990-2024. This yields 1,021 country-year observations with full rating coverage. Pension spending data (from OECD social expenditure database) covers 815 observations; sovereign spreads cover 689 observations.

### 2.2 Key Variables

| Variable | Source | Coverage |
|:---|:---|---:|
| Sovereign rating (21-point numeric) | S&P via Kose et al. | 1,021 obs |
| Pension spending/GDP | OECD SOCX | 815 obs |
| Health expenditure/GDP | World Bank WDI | 668 obs |
| Government expenditure/GDP | IMF WEO | 923 obs |
| Government revenue/GDP | IMF WEO | 923 obs |
| Expenditure-revenue gap | Computed | 923 obs |
| Government debt/GDP | IMF WEO | 909 obs |
| Sovereign spread (lending - govt) | WDI / IFS | 689 obs |
| Old-age dependency ratio | UN WPP | 1,021 obs |

### 2.3 Estimation

All specifications use PanelGLS with AR(1) error correction (Prais-Winsten), following the methodology used throughout the research program. Standard controls include log GDP per capita, GDP growth, and inflation. Country and year fixed effects are absorbed by the within-country variation in the GLS transformation.

---

## 3. The Pension Endogeneity Channel

### 3.1 Aging Drives Expenditure, Not Revenue

**[Table 1: Pension Endogeneity]** (see `output/tables/table1_pension_endogeneity.md`)

The first stage of the doom loop is strongly supported. Old-age dependency predicts:

- **Pension spending/GDP**: coefficient 50.7 (p<0.001). A 10 percentage point increase in OADR raises pension spending by 5.1 percentage points of GDP.
- **Health expenditure/GDP**: coefficient 16.0 (p<0.001). Healthcare costs respond to aging, but at one-third the rate of pensions.
- **Government expenditure/GDP**: coefficient 50.8 (p<0.001). Nearly all of the expenditure response is pensions.
- **Government revenue/GDP**: coefficient 8.4 (p=0.45). Revenue does not respond to aging. This is the expenditure-revenue asymmetry documented in the fiscal dominance paper, confirmed in the sovereign-rated issuer sample.
- **Expenditure-revenue gap**: coefficient 20.6 (p<0.001). The gap widens mechanistically.
- **Government debt/GDP**: coefficient 290.8 (p<0.001). The cumulative effect on debt is massive.

The asymmetry is striking: the expenditure coefficient (50.8) is six times the revenue coefficient (8.4). This exceeds the 3.3:1 ratio from the fiscal dominance paper, likely because the sovereign-rated sample is more advanced (higher baseline pension systems). The revenue null (p=0.45) means that aging economies cannot grow their way out of pension obligations through the tax base — the fiscal adjustment must come from spending restraint or debt accumulation.

### 3.2 The Median Voter Mechanism

The pension endogeneity result is consistent with the median voter theory of fiscal generosity. As the electorate ages, the median voter is older, increasing political demand for pension generosity. This creates a ratchet effect: pension promises made during demographic expansion are difficult to retract as the dependency ratio rises.

---

## 4. The Fiscal-Pension Feedback

### 4.1 Pension Spending Interacts with Demographics to Accelerate Debt

**[Table 2: Fiscal-Pension Feedback]** (see `output/tables/table2_fiscal_pension_feedback.md`)

The second link in the doom loop is the fiscal-pension feedback. Pension spending alone does not predict debt changes (p=0.44) — many countries successfully manage pension costs through reforms, indexation rules, and productivity growth. However, the interaction of pension spending with old-age dependency is significant (p=0.04): in aging economies, pension spending accelerates debt accumulation disproportionately.

The expenditure-revenue gap provides an even cleaner channel. The gap directly predicts debt changes (0.267, p<0.001), and the gap-OADR interaction amplifies this (2.38, p<0.001). Countries with both high expenditure gaps and aging populations accumulate debt faster than either factor alone would predict.

This nonlinearity explains why some high-pension countries (e.g., Denmark, Sweden) manage their fiscal positions while others (e.g., Italy, France) do not: the outcome depends on the interaction with the demographic trajectory, not on pension levels alone.

---

## 5. Sovereign Risk Amplification

### 5.1 The Doom Loop Test

**[Table 3: Sovereign Risk Amplification]** (see `output/tables/table3_doom_loop_amplification.md`)

The third link connects the fiscal deterioration to sovereign risk. Pension spending directly predicts sovereign ratings (p=0.03): higher pension spending is associated with lower ratings, even controlling for debt levels, growth, and income.

The OADR x debt interaction on ratings is marginally significant (p=0.07). This replicates the safe asset cliff paper's finding in a different specification: high debt is more damaging to ratings in aging economies. The effect is consistent with the spline at 20% OADR documented in Paper 16 (-17.49***).

The triple interaction (OADR x debt x expenditure gap) is not significant (p=0.81), indicating that the doom loop operates through pairwise channels rather than a single three-way mechanism. This is empirically reasonable: the doom loop is a sequential chain (aging → pension → debt → rating), not a simultaneous three-way interaction.

### 5.2 Spread Dynamics

Sovereign spread results are weaker, with the OADR x debt interaction on spreads not significant (p=0.61). This is consistent with the regulatory overlay hypothesis: zero risk weights suppress spread signals until a rating threshold is crossed, compressing the identifying variation.

---

## 6. Dynamic Propagation

### 6.1 Local Projections

**[Table 4: Local Projections]** (see `output/tables/table4_local_projections.md`)

We test dynamic propagation using local projections at horizons h=1,2,3,5. The pension-OADR interaction on cumulative rating changes is directionally negative (shocks worsen ratings) at all horizons but not statistically significant. The expenditure gap-OADR interaction on cumulative debt changes is also insignificant.

The null on dynamic propagation likely reflects power limitations: 31 countries with annual data and sparse pension coverage (815 of 1,021 observations) provide insufficient variation for multi-horizon identification. The cross-sectional evidence (Phases 2-4) is better powered because it uses the full within-country variation over 35 years.

This null should be interpreted as an absence of evidence rather than evidence of absence. Quarterly data on sovereign spreads, ratings actions, and institutional holdings would provide better identification of the doom loop's dynamic propagation.

---

## 7. Country-Level Doom Loop Intensity

### 7.1 Current Rankings

**[Table 5: Doom Loop Intensity by Country]** (see `output/tables/table5_doom_loop_intensity.md`)

We construct a composite doom loop score for each country based on four normalized components: old-age dependency (35%), debt/GDP (35%), expenditure-revenue gap (30%). Countries are ranked by their current doom loop intensity.

The top five most vulnerable countries are:

1. **Japan** (score 0.80): Highest OADR and highest debt in the sample. No longer rated as safe issuer (downgraded from AAA in 2001, currently AA-). The doom loop is already advanced.

2. **France** (score 0.73): Still a safe issuer (AA) but with rising OADR, 113% debt/GDP, and a 5.8pp expenditure-revenue gap. Pension spending at 30.6% of GDP is among the highest in the sample.

3. **Italy** (score 0.73): Lost safe-issuer status. 135% debt/GDP with pension spending at 27.6%. The doom loop has progressed past the rating cliff.

4. **Finland** (score 0.70): Still a safe issuer (AA+) but rapidly aging. Pension spending at 31.4% is the highest in the sample.

5. **Belgium** (score 0.66): Safe issuer (AA-) with 104% debt/GDP and high pension spending (28.6%).

### 7.2 The Safe Issuer Paradox

Six of the top ten doom loop countries are currently rated as safe issuers (AA- or above): France, Finland, Belgium, Austria, Germany, and the United States. These countries hold zero risk weights under Basel III, meaning that banks and insurers face no capital charge for holding their sovereign debt. This creates the regulatory lock-in: institutional investors are incentivized to hold precisely the sovereigns most vulnerable to the doom loop.

The safe asset cliff paper projects that 10 of 24 current safe issuers will lose that status by 2054. Our doom loop analysis identifies the mechanism: the pension-debt feedback is most intense in safe issuers because they are the most demographically advanced.

### 7.3 Projections

**[Table 6: Doom Loop Projections]** (see `output/tables/table6_doom_loop_projections.md`)

Using UN demographic projections and current fiscal stances, we project doom loop intensity through 2050. The projections assume that fiscal policy parameters (expenditure-revenue gap, debt trajectory) remain at current levels while demographics follow UN medium-variant projections.

Key findings:
- Korea, Spain, and Poland are projected to experience the largest increases in doom loop intensity, as their OADR rises rapidly from currently moderate levels.
- Japan's doom loop intensity remains the highest throughout the projection period but increases more slowly (already near peak aging).
- Germany and Austria face rising intensity despite current fiscal prudence, as the demographic denominator worsens.

---

## 8. The Regulatory Overlay

### 8.1 Basel III and Solvency II

This section provides a qualitative assessment of the regulatory overlay, as direct data on institutional sovereign holdings by risk-weight category is not available in our panel.

Under Basel III, sovereign debt rated AA- or above carries a 0% risk weight for credit risk. Solvency II applies similar treatment for insurance companies. This means that pension funds, insurance companies, and banks can hold unlimited amounts of safe-rated sovereign debt without setting aside any regulatory capital.

The regulatory incentive is clear: holding safe sovereign debt is "free" from a capital perspective, while diversifying into other assets incurs capital charges. In aging economies where pension funds are the largest institutional investors, this creates a natural demand for domestic sovereign debt that is amplified by home bias.

### 8.2 The Rating Cliff Problem

The doom loop's danger lies not in the gradual deterioration of sovereign creditworthiness but in the *discontinuity* at the rating cliff. When a sovereign is downgraded below AA- (the safe threshold), risk weights jump from 0% to 20%, immediately increasing capital requirements for all institutional holders. This triggers forced selling, spread widening, and potentially a crisis — the kind of abrupt adjustment that the doom loop's gradual fiscal erosion makes increasingly likely.

The safe asset cliff paper projects that the median safe supply ratio declines from 89% (2024) to 77% (2054), with the 10th percentile reaching 13.5% — implying a scenario where most current safe issuers have lost that status. Our doom loop analysis identifies the mechanism driving this projection: it is not exogenous shocks but the endogenous fiscal deterioration driven by pension obligations in aging economies.

### 8.3 Policy Implications

The regulatory doom loop suggests three policy interventions:

1. **Demographic stress testing for sovereign risk weights**: Basel III risk weights should incorporate forward-looking demographic indicators (OADR projections, pension spending trajectories) rather than relying solely on backward-looking credit ratings.

2. **Pension mandate reform**: Mandatory sovereign allocation rules for pension funds should be relaxed or eliminated, allowing funds to diversify away from domestic sovereigns whose fiscal backing is demographically compromised.

3. **Gradual risk weight adjustment**: Rather than maintaining the 0%/20% cliff at the AA- threshold, risk weights should increase gradually with demographic-fiscal indicators, providing earlier market signals and preventing the abrupt adjustment at the rating cliff.

---

## 9. Limitations

### 9.1 Data Limitations

Our panel covers only 31 sovereign-rated issuers with pension data for 815 of 1,021 observations. This limits statistical power, particularly for interaction terms and local projections. The null on dynamic propagation (Phase 5) may reflect insufficient power rather than absent dynamics.

### 9.2 Absence of Institutional Holdings Data

We cannot directly test the regulatory overlay mechanism because we lack panel data on institutional sovereign holdings broken down by risk-weight category. ECB MFI data, BIS consolidated banking statistics, and Solvency II reporting data would enable direct testing of whether institutional holdings concentrate in doom-loop-vulnerable sovereigns. This is a priority for future work.

### 9.3 Endogeneity

Pension spending and sovereign ratings are both endogenous to the same macroeconomic environment. Our PanelGLS specification mitigates but does not eliminate this concern. The predetermined demographic variable (OADR projected 20 years forward) provides a quasi-exogenous instrument for the first stage (aging → pension spending), but the second and third stages (pension → debt → rating) are more vulnerable to reverse causality.

---

## 10. Conclusion

The demographic regulatory doom loop is a slow-moving structural risk to sovereign creditworthiness in advanced economies. The mechanism is well-identified in cross-section: aging drives pension spending (6:1 faster than revenue), pension spending interacts with demographics to accelerate debt, and the debt-demographics interaction predicts rating deterioration. The regulatory overlay — zero risk weights for safe sovereigns — suppresses market signals and forces institutional exposure to precisely the countries most vulnerable to this loop.

The policy implication is urgent: by the time credit ratings reflect the demographic fiscal erosion, the regulatory cliff effect will make the adjustment abrupt. Incorporating forward-looking demographic indicators into sovereign risk assessments — whether through stress testing, risk weight reform, or pension mandate liberalization — would provide earlier signals and smoother adjustment paths.

The countries most at risk — Japan, France, Italy, Finland, Belgium — are among the world's most systemically important sovereign debtors. Their doom loop trajectories are not tail risks; they are the central forecast under current demographic and fiscal trends.

---

## References

Bohn, H. (1998). The behavior of US public debt and deficits. *Quarterly Journal of Economics*, 113(3), 949-963.

De Nardi, M., Imrohoroglu, S. and Sargent, T. (2016). Savings and demographic change: the global dimension. *Unpublished manuscript*.

Peters, B. (2026). Population Aging and the Fiscal Sustainability Trap: Expenditure Asymmetry, Debt Dynamics, and the Limits of the Interest Rate Channel. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). The Safe Asset Cliff: Demographic Downgrade Risk and Collateral Scarcity. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Sample Composition Fragility in Demographic-Macro Research: A Diagnostic Framework. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). When Does Demography Move Capital? A Nonlinear Framework for Conditional Demographic Effects. *Perspective on Risk Working Paper Series*.

---

### Companion Papers in This Series

Peters, B. (2026). Demographic Structure and International Capital Flows: Evidence from 140 Countries. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Where Does Demographic Capital Go? Bilateral Evidence from a Gravity Model. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Does Demography Cause Capital Flows? Evidence from Post-Soviet Transitions. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Does Demographic Capital Do Anything? Capital Deepening, the Allocation Puzzle, and the J-Curve of Demographic Investment. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Demographics and Asset Prices: The Murder-Suicide of the Rentier. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Demographics and Japanification: Who's Next? *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Population Aging and the Fiscal Sustainability Trap. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Demographics and Financial Crises: Age Structure as an Early Warning Signal. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). The Demographic Erosion of Fiscal Leverage: Twin Deficits in an Aging World. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Why Feldstein-Horioka Correlations Vary: Demographics and the Savings Retention Puzzle. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Demographics and the Trilemma: How Population Aging Shapes Exchange Rate Regime Choice. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Demographics, Investment, and Capital Deepening: How Aging Reshapes Production Structure. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Net vs Gross External Adjustment: Demographics as a Latent Factor. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Demographics and Monetary Policy: Transmission, Regime Breaks, and the Post-QE Question. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Demographic Divergence and Safe Asset Scarcity. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). The CCA Tipping Point: Post-Soviet Transitions as a Natural Experiment. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). The Safe Asset Cliff: Demographic Downgrade Risk and Collateral Scarcity. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). When Does Demography Move Capital? A Nonlinear Framework for Conditional Demographic Effects. *Perspective on Risk Working Paper Series*.

Peters, B. (2026). Sample Composition Fragility in Demographic-Macro Research: A Diagnostic Framework. *Perspective on Risk Working Paper Series*.

## Appendix: Output Tables

- Table 1: Pension Endogeneity (`table1_pension_endogeneity.md`)
- Table 2: Fiscal-Pension Feedback (`table2_fiscal_pension_feedback.md`)
- Table 3: Sovereign Risk Amplification (`table3_doom_loop_amplification.md`)
- Table 4: Local Projections (`table4_local_projections.md`)
- Table 5: Doom Loop Intensity by Country (`table5_doom_loop_intensity.md`)
- Table 6: Doom Loop Projections (`table6_doom_loop_projections.md`)
